Residual Coupling (RC) connects frozen language models in parallel using lightweight learned linear bridges, enabling horizontal scaling without weight modification. It reduces perplexity by up to 80.7% compared to MoE and improves accuracy on TruthfulQA by 9.1 percentage points.
Residual Coupling (RC) connects frozen language models in parallel using small, learned linear bridge projections. These bridges read hidden states from one model and inject additive updates into the residual stream of another at intermediate layers. In bilateral setups, simultaneous return bridges form a feedback loop that stabilizes both streams without altering base weights. This architecture establishes a two-step paradigm where base models function as memorizers, while lightweight linear bridges handle cross-domain generalization. Constraining the bridges to purely linear maps prevents overfitting because they can only map existing geometric relationships between the frozen representation spaces. As the bridges are optimized against ground-truth target data, they have no incentive to map ungrounded features such as individual models' hallucinations. Keeping the base weights completely frozen eliminates catastrophic forgetting. The system maintains operational closure, transforming inputs through its existing structure rather than changing to accommodate them. Evaluating bilateral RC against Mixture-of-Experts (MoE) routing across the same frozen models shows these results: * Medical (3-model): Reduces perplexity to 11.02, compared to 56.80 for MoE and 57.08 for the frozen baseline. This represents an 80.7% reduction. * TruthfulQA Health (MC1): Improves accuracy by 9.1 percentage points over the baseline. Independent models have uncorrelated hallucinations, allowing the bridge gates to amplify consistent cross-model updates while suppressing individual errors. * Coding Test: CodeGPT-small-py and GPT-2 use different tokenizers, causing a 7-million baseline perplexity on mismatched text. MoE reaches 878, but RC achieves 5.91 by reading hidden states before the output projection collapses. This framework introduces a horizontal scaling axis for multi-model systems, moving beyond vertical scaling via larger monolithic models. Latency remains bounded by the slowest single model. Specialists can be added or removed without retraining the remaining system. In some scenarios, this architecture could replace multi-turn text prompting in agentic workflows with a single parallel forward pass, allowing models and/or bridges to run on separate nodes or edge devices without a central bottleneck. By decoupling memorization from relational alignment, RC bridges provide a framework for scaling multi-model systems and offer a path toward native multi-modal integration. Paper: [https://ssrn.com/abstract=6746521](https://ssrn.com/abstract=6746521) Code: [https://github.com/pfekin/residual-coupling/](https://github.com/pfekin/residual-coupling/)
This paper introduces Hidden Decoding, a sequence-length scaling method for LLMs that adds internal computation per token by expanding each token into multiple streams with independent embeddings, using Stream-Factorized Attention to keep costs low. Experiments on models up to 617B parameters show consistent improvements over baselines, demonstrating a practical fixed-backbone scaling path.
R²-dLLM introduces spatio-temporal redundancy reduction techniques that cut diffusion LLM decoding steps by up to 75% while preserving generation quality, addressing a key deployment bottleneck.
This paper introduces RLearner-LLM, a framework using Hybrid-DPO to balance logical correctness and fluency in LLM-generated explanations, achieving significant NLI entailment improvements across multiple domains and base models while mitigating the verbosity bias of standard preference signals.
OmniLens is a scalable lens method for interpreting LLM hidden states, using low-rank translators and Subset-KL to reduce parameters and memory, enabling a dense ensemble of 482 lenses on LLaMA-3.3-70B at substantially lower cost.
Introduces cMoLLM, a convolutionally gated mixture-of-LLMs that routes over end-to-end streams through fully differentiable dynamic convolution, improving perplexity and downstream accuracy under matched compute.